Repeating certain part of string conditionally

Question

I want to repeat certain part of string between ] and ; as the number of elements separated by ; preceding within []. So the desired output for [A1, AB11; A2, AB22] I1, C1 would be [A1, AB11] I1, C1; [A2, AB22] I1, C1. Any hints to start with. Thanks

df1 <-
  data.frame(
   String = c(
    "[A1, AB11; A2, AB22] I1, C1; [A3, AB33] I3, C1"
  , "[A4, AB44] I4, C4; [A5, AB55; A6, AB66; A7, AB77] I7, C7"
  )
  )
df1

                                                    String
1           [A1, AB11; A2, AB22] I1, C1; [A3, AB33] I3, C1
2 [A4, AB44] I4, C4; [A5, AB55; A6, AB66; A7, AB77] I7, C7


df2 <-
  data.frame(
   String = c(
    "[A1, AB11] I1, C1; [A2, AB22] I1, C1; [A3, AB33] I3, C1"
  , "[A4, AB44] I4, C4; [A5, AB55] I7, C7;[A6, AB66] I7, C7; [A7, AB77] I7, C7"
  )
  )

df2

                                                                     String
1                   [A1, AB11] I1, C1; [A2, AB22] I1, C1; [A3, AB33] I3, C1
2 [A4, AB44] I4, C4; [A5, AB55] I7, C7;[A6, AB66] I7, C7; [A7, AB77] I7, C7

how much times you want to repeat them – Abdessabour Mtk Nov 24 '20 at 19:07 — Abdessabour Mtk, Nov 24 '20 at 19:07
As the number of elements separated by `;` within `[]`. – MYaseen208 Nov 24 '20 at 19:11 — MYaseen208, Nov 24 '20 at 19:11
ok now I get the question – Abdessabour Mtk Nov 24 '20 at 19:12 — Abdessabour Mtk, Nov 24 '20 at 19:12

score 2 · Answer 1 · answered Nov 24 '20 at 19:14

Here's a base R solution:

sapply(strsplit(paste0(df1$String, ";"), "\\[|\\]"), function(x) {
  for(i in seq_along(x))
  {
    if(i %% 2 == 0) {
      x[i] <- paste0("[", gsub(";", paste0("]", x[i + 1], " ["), x[i]), "]")
    }
  }
  paste(x, collapse = "")
})
#> [1] "[A1, AB11] I1, C1;  [ A2, AB22] I1, C1; [A3, AB33] I3, C1;"                   
#> [2] "[A4, AB44] I4, C4; [A5, AB55] I7, C7; [ A6, AB66] I7, C7; [ A7, AB77] I7, C7;"

score 2 · Answer 2 · answered Nov 24 '20 at 21:59

I had tried something similar in the past, and thought it might be interesting to adapt using glue and unglue packages.

The initial strsplit separates by semicolon, ignoring the semicolons between the brackets.

The unglue will separate out what is between brackets that gets repeated, and what is appended outside the brackets, for each row.

library(glue)
library(unglue)
library(purrr)

my_fun <- function(inside, outside) {
  glue("[{inside}] {outside}")
}

sapply(strsplit(df1$String, '\\[[^]]*\\](*SKIP)(*F)|;\\s', perl = T), function(x) {
  ud <- unglue_data(x, patterns = "[{Inside}] {Outside}")
  ud_in <- map(ud[['Inside']], strsplit, split = "; ")
  ud_map <- map(seq_along(ud[['Inside']]), function(y) {
    map2(unlist(ud_in[y]), ud[['Outside']][y], my_fun)
  })
  paste(unlist(ud_map), collapse = '; ')
})

Output

[1] "[A1, AB11] I1, C1; [A2, AB22] I1, C1; [A3, AB33] I3, C1"                   
[2] "[A4, AB44] I4, C4; [A5, AB55] I7, C7; [A6, AB66] I7, C7; [A7, AB77] I7, C7"

Abdessabour Mtk · Accepted Answer · 2020-11-24T20:03:48.190

Not the tidiest solution however it's using stringr

str_split(df1$String, ";(?= *\\[)") %>%
  map(str_match, "\\[(.+?)\\] (.+)") %>%
   map( ~ paste(unlist(map2(paste0(str_split(.x[,2], "; ?")), .x[,3], ~ paste0("[", .x,"] ",.y ))), collapse="; "))

somewhat nicer solution:

as_tibble(df1) %>%
  mutate(splits=str_split(String, "; *(?=\\[)")) %>%
   unnest_longer(col=splits) %>%
    mutate(splits=map(str_split(splits,"\\[|\\] ?"), str_split, "; ?"))  %>%
     unnest_wider(splits) %>%
      mutate(val=map2(...2, ...3, ~ paste0("[", .x ,"] ", .y, collapse="; ") )) %>%
       group_by(String) %>%
        summarise(val=paste0(val, collapse="; "))
# A tibble: 2 x 2
  String                             val                                        
  <fct>                              <chr>
1 [A1, AB11; A2, AB22] I1, C1; [A3,… [A1, AB11] I1, C1; [A2, AB22] I1, C1; [A3, AB33] I3, C1
2 [A4, AB44] I4, C4; [A5, AB55; A6,… [A4, AB44] I4, C4; [A5, AB55] I7, C7; [A6, AB66] I7, C7; [A7, AB77] I7, C7

Repeating certain part of string conditionally

3 Answers3